Fresh data and how-to advice to lift per­for­mance and get more con­ver­sions

Ear­li­er this year, Face­book broke some bad news. Organ­ic reach is offi­cial­ly being choked, mak­ing it hard­er for brands to reach the audi­ences they’ve worked so hard to build.

Because of this, I believe mar­keters will look to SEM (search engine mar­ket­ing) to recap­ture lost atten­tion. The prob­lem is, there’s already so much com­pe­ti­tion. How do you get past the noise and gen­er­ate PPC (pay per click) results, and which KPIs (key per­for­mance indi­ca­tors) should you be track­ing to mea­sure suc­cess?

Opti­miz­ing con­ver­sion rate (CVR) is one of the fastest ways to improve AdWords effi­cien­cy. It allows you to test new approach­es and boost ROI with­out hav­ing to expand tar­get key­words, cam­paigns or bud­get.

Here are five approach­es to PPC that will help you gen­er­ate more con­ver­sions and bet­ter results in 2018 and beyond.

1. Optimize keyword quality score

Google’s entire busi­ness mod­el relies on pro­vid­ing searchers with rel­e­vant results. This goes for organ­ic results as well as AdWords.

To do this, Google assigns your tar­get key­words a Qual­i­ty Score (QS). This QS, along with your CPC (cost per click) bid, is what then deter­mines your “Ad Rank.”

The three ele­ments that deter­mine your QS are:

  • Ad rel­e­vance (in oth­er words, how rel­e­vant the key­word is to the ad copy you serve).
  • Land­ing page expe­ri­ence.
  • Expect­ed CTR (click-through rate).

Many PPC experts con­sid­er CTR the most impor­tant fac­tor when deter­min­ing QS. There­fore, when look­ing to opti­mize your QS, start with CTR.

Ana­lyze the key­word rel­e­van­cy of your cam­paigns. Is your ad copy aligned with the search intent of the key­word?

It’s good prac­tice to cre­ate sep­a­rate Ad Groups for each of your key­words. Also known as Sin­gle Key­word Ad Groups, this is where you cater to the intent of spe­cif­ic searchers rather than a larg­er group.

In the exam­ple below (cour­tesy of Con­ver­sionXL), ASDA is the only adver­tis­er for the term “wom­ens red dress­es” with copy tai­lored to that search term.

example PPC

As well as rel­e­van­cy, your ad copy should quick­ly sell the ben­e­fits of the “click.” In oth­er words, why should the searcher pay atten­tion? Make your head­lines rel­e­vant, focus­ing on the desires and pain points of your audi­ence.

By opti­miz­ing CTR, and there­fore qual­i­ty score, you’ll gen­er­ate more qual­i­fied traf­fic. And high-qual­i­ty traf­fic deliv­ers bet­ter con­ver­sion rates.

Once you’ve opti­mized CTR, your land­ing pages should be the next tar­get. Dynam­ic text replace­ment (DTR) can pro­vide some quick wins. This “swaps” spe­cif­ic copy in your land­ing page based on the key­word the user searched to find you. DTR can improve qual­i­ty score and there­fore con­tribute to a high­er CVR.

2. Intelligent remarketing

When it comes to AdWords, high bounce rates are a fact of life. Users who come to your land­ing pages are at var­i­ous stages of the cus­tomer jour­ney. For exam­ple, a call-to-action (CTA) for a demo won’t work on a searcher who is still edu­cat­ing them­selves on dif­fer­ent solu­tions.

To cap­ture these missed oppor­tu­ni­ties, use remar­ket­ing to cross-sell and “down-sell” bounced vis­i­tors. Let’s start by expand­ing on the exam­ple above. If you’re offer­ing a demo of your soft­ware to some­one who is still in the aware­ness phase, this approach won’t be as effec­tive as some­thing that answers their ques­tions.

There­fore, an e‑book that teach­es prospects how to over­come spe­cif­ic chal­lenges is an appro­pri­ate down-sell. It would edu­cate them on the options avail­able to them while pro­vid­ing infor­ma­tion about how your prod­uct makes the process eas­i­er.

Of course, these chal­lenges will vary depend­ing on per­sonas and cus­tomer seg­ments. There­fore, you must per­son­al­ize your ad cre­ative where nec­es­sary.

Retar­get­ing in this way allows you to cap­ture lead infor­ma­tion that would have been oth­er­wise lost, boost­ing the CVR and over­all ROI of your cam­paigns. The mis­take many mar­keters make here is to “re-sell” the demo request. Use it as an oppor­tu­ni­ty to edu­cate them and add more val­ue instead of forc­ing them fur­ther down the fun­nel.

Here are some tips you can apply to your remar­ket­ing ads to cap­ture the atten­tion of lost leads:

  • Test dif­fer­ent lead mag­nets: Dif­fer­ent per­sonas and cus­tomer types respond to dif­fer­ent forms of media. Split-test your remar­ket­ing ads to offer an e‑book and webi­nar. See which gen­er­ates the high­est con­ver­sions and dou­ble down on those for­mats.
  • Name-drop influ­encers: If you work with well-known influ­encers in your space, con­sid­er includ­ing them in your remar­ket­ing ads. This asso­ci­a­tion adds an ele­ment of trust like no oth­er.
  • Use dynam­ic tar­get­ing: Serve spe­cif­ic ads to dif­fer­ent audi­ence seg­ments. More on this lat­er.

The point of remar­ket­ing is to cap­ture lost users and retain cus­tomers. Don’t waste the oppor­tu­ni­ty by serv­ing the same mes­sag­ing. Look for ways to add val­ue up and down the fun­nel.

3. Tap into the power of machine learning

AI and machine learn­ing bring the promise of high­er-per­form­ing mar­ket­ing at speed. From an AdWords per­spec­tive, this would mean auto­mat­ed bid and bud­get man­age­ment, using more data than a human can han­dle to make adjust­ments in real time.

To find out exact­ly what impact machine learn­ing has on PPC per­for­mance, we ana­lyzed 32,858 paid accounts using the Acquisio Tur­ing plat­form to uncov­er the truth. Here’s what we found out about con­ver­sions and machine learn­ing:

  1. An aver­age increase in con­ver­sions of 71 per­cent.
  2. A medi­an increase in con­ver­sions of 22 per­cent.

Dis­cus­sions of land­ing page qual­i­ty aside, the huge dif­fer­ence between aver­age and medi­an is explained by the fact that a cer­tain num­ber of accounts saw extreme­ly high increas­es in num­ber of con­ver­sions, which skews the aver­age in a mean­ing­ful way. If we wished to exclude those extremes from the dis­cus­sion, we would look at the medi­an score, which tells us the per­cent increase in con­ver­sions that was observed for the 50th per­centile.

The plot thick­ened because this increase in con­ver­sions came with an over­all decrease in cost per acqui­si­tion (CPA). In fact, the medi­an CPA had a decrease of 18 per­cent, with 64 per­cent of the group enjoy­ing a decrease in CPA over­all.

While the report above focused on the increase in con­ver­sions made pos­si­ble by machine learn­ing, our most recent study exam­ined 50,000 cam­paigns to deter­mine Google AdWords Indus­try Bench­marks and looked at con­ver­sion rate (CVR) with and with­out machine learn­ing by indus­try. Here are the CVR find­ings seg­ment­ed by busi­ness cat­e­go­ry:

Con­ver­sion rate (CVR) by indus­try with and with­out machine learn­ing

conversion rate

Machine learn­ing martech helps PPC mar­keters scale and opti­mize mar­ket­ing activ­i­ties effi­cient­ly, but it’s also a seri­ous con­tender for con­ver­sion boosts.

Here’s the thing: Machine learn­ing tech­nolo­gies get bet­ter the more they learn. In oth­er words, results will improve as machine learn­ing algo­rithms react to new find­ings. Check out The Marketer’s Field Guide to Machine Learn­ing for more infor­ma­tion.

4. Test new ad extensions

To cut through the noise, you must cap­ture as much SERP (search engine results page) real estate as pos­si­ble. This means not only stand­ing out with your cre­ative but also expand­ing how much room your ads take up.

To do this, test dif­fer­ent ad exten­sions on your top-per­form­ing cam­paigns. Ad exten­sions, as defined by Google, “expand your ad with addi­tion­al infor­ma­tion — giv­ing peo­ple more rea­sons to choose your busi­ness. They typ­i­cal­ly increase an ad’s click-through rate by sev­er­al per­cent­age points.”

Ad exten­sions come in sev­er­al forms, the most pop­u­lar of which are:

  • Sitelink Exten­sions: Pro­vide links to oth­er rel­e­vant pages on your web­site.
  • Call­out Exten­sions: Addi­tion­al infor­ma­tion on what you’re offer­ing, e.g., lim­it­ed stock and free deliv­ery.
  • Struc­tured Snip­pets: Allows you to high­light spe­cif­ic ele­ments. For exam­ple, if you’re sell­ing “Ital­ian veg­an leather boots,” you can include a list of shoe sizes.
  • Loca­tion Exten­sions: Include your busi­ness address and tele­phone num­ber in your ad copy.

As you’re well aware, mobile user behav­ior is very dif­fer­ent from desk­top users’. Indeed, 61.9 per­cent of all PPC clicks were from a smart­phone dur­ing Q3 of 2017.

Google has react­ed to this shift in behav­ior by adding addi­tion­al exten­sions for ads that appear on mobile devices. These are:

  • Mes­sage Exten­sions: Allow users to send an SMS to your busi­ness direct­ly from the SERPs.
  • Call Exten­sions: Sim­i­lar­ly, users can dial a phone num­ber pro­vid­ed with­in your ad copy.

As always, test dif­fer­ent exten­sions on a small scale before apply­ing them to all of your cam­paigns. Keep the customer’s jour­ney and intent in mind. Are they search­ing for a term with sev­er­al pos­si­ble out­comes? Con­sid­er using a Sitelink exten­sion. Does it look like they’re search­ing for your retail store on a mobile phone? Include mobile exten­sions.

5. Advanced segmentation with in-market audiences

Face­book Ads are pop­u­lar among mar­keters due to the advanced tar­get­ing avail­able. But many are still unaware of AdWords’ func­tion­al­i­ty to do the same.

Google col­lects a tremen­dous amount of data on their users. So it was only a mat­ter of time before they allowed mar­keters to use it them­selves.

That’s where in-mar­ket audi­ences come in. By using in-mar­ket audi­ences with­in your dis­play ads tar­get­ing, you can tar­get users based on their con­sumer behav­ior, as well as the con­tent they have expressed an inter­est in online.

The data avail­able is sort­ed into sev­er­al mar­ket cat­e­gories, includ­ing real estate, trav­el and telecom­mu­ni­ca­tion. You can then set tar­get­ing on a gran­u­lar lev­el, all the way down to spe­cif­ic inter­ests and brand names:


So, how does it work? Accord­ing to Google, data such as sites browsed, the prox­im­i­ty of vis­its, rel­e­vant ads clicked and con­ver­sions are all tak­en into account to cat­e­go­rize users by intent.

This means that, while this is lim­it­ed to the Dis­play net­work only, you’re able to serve hyper-spe­cif­ic ads to those who have expressed an inter­est. From per­sona seg­ments to prod­uct cat­e­gories, the options are many.